The goal of this proposal is to lay the foundation for the next generation of Peer Change Agent (PCA) HIV prevention interventions. Our approach is guided by identification of candidate PCAs according to their structural position within a social network and complements traditional approaches that utilize ethnographic observation or expert opinion. This intervention approach could dramatically reduce HIV incidence and bring us closer to HIV elimination. During the past 20 years, the most widely used PCA network based intervention for HIV prevention has been the opinion leader (OL) model or variants of it. These approaches and other PCA interventions include training popular members or peers from a target population to promote HIV prevention messages through interpersonal communication. Attempts to implement these models in diverse settings, however, have been disappointing. Fundamental questions, about the selection criteria for PCAs exist given the heterogeneity of approaches utilized and criteria for selection, has yet to be empirically tested. The effectiveness of the OL model can be further limited by several factors, including inadequate network assessments, and the OL's public position which often requires she/he maintain conformity to the status quo, thus potentially limiting innovation adoption. Consequently, identifying those who connect across groups (hereafter bridging actors), may be an alternative/complementary and efficient way to use network information to accelerate behavior change as has been demonstrated in workplace and management settings. In order to develop next generation PCA interventions, we approach gaps in social network data collection, network analysis and systematic PCA selection through the following Specific Aims: 1) Characterize a communication network through an approach that rapidly links Facebook friend lists of YBMSM on the South Side of Chicago. The communication network will be characterized as in recent work such that study respondents (n=300) rapidly generate a large linked MSM network (n=3500); 2) Measure attributes of individuals and dyads within the MSM network including sociodemographics leadership/innovation potential, condom communication, and tie qualities. These individual and dyadic attributes are believed to be important features of PCAs and will be determined by self-report and by report from others in the network; 3) Explore processes hypothesized to underlie behavior change in PCA interventions among both centrally located (popular) individuals and those who are in bridging positions (link distinct groups) within the network generated in Aim 1. We will identify candidate PCAs (n=30) who are selected based upon their network position and conduct in-depth qualitative interviews to assess candidate PCA behavior change elements that are considered to be important to PCA based interventions: education, social support, social norms, self-efficacy and advocacy.